I was reading an ACM research paper recently that considered users' revisitation patterns. The authors found that users tend to revisit native apps far more often than they revisit pages in browsers.

The clear message to web designers is to point users to installing the NIA to increase the probability of revisits (2775).

Yet, there is a growing popular conception, as evidenced by a recent xkcd comic, that native apps are of lower quality than in-browser mobile sites and apps.

xkcd comic Source: xkcd: App

As the comic indicates, it's annoying to users to prompt them repeatedly to download the app. Yet the research paper indicates that it's in your best interest to nudge users toward downloading because doing so will increase their likelihood of revisitation.

Let's assume that you have created an app of equal or higher quality to the mobile site. Let's also assume that some of your users are tired of downloading apps of lower quality than the web-app or mobile site versions.
Without annoying users, how do you nudge them toward downloading your app? Should you even be trying to nudge them to download in the first place?

  • There's a interesting post on the MailChimp blog which details how they marketed their Gather app to customers by segmenting them into different groups. It's email marketing, but I think it has a strong principle: only market to those who are interested in what you are offering.
    – Brendon
    Commented Apr 13, 2013 at 23:07
  • 1
    Did the paper state why native apps are revisited more?
    – Dvir Adler
    Commented Apr 14, 2013 at 6:22
  • @DvirAdler They suggested that it was because native apps were being used as a form of bookmarking for smartphones in lieu of traditional bookmarking in a browser: users installed NIAs [native internet applications] that “stuck” to their vocabularies for longer periods of time. In a sense, this emphasis on a new type of bookmarking (installing NIAs to springboards) has afforded iPhone users to optimize their devices. Commented Apr 14, 2013 at 15:22
  • So isn't there a way to make the phone browser add the a shortcut to the website to the springboard (so it just looks like a native app, but it is only a shortcut)?
    – Dvir Adler
    Commented Apr 18, 2013 at 9:15
  • it's annoying to users to prompt them repeatedly to give you all their money. Yet basic finance indicates that it's in your best interest to nudge them to give you all their money. So should you do it?
    – Juan Perez
    Commented Dec 15, 2021 at 12:38

5 Answers 5


The primary things that are most annoying (at least for me) with these banners are:

  1. I generally get the notifications even if I have the app already.

  2. I get notifications for sites that don't warrant their own apps in the first place (the main culprit here is discussion forum software, which seems to come bundled with nagware nowadays), or at best the site doesn't justify the benefits I'd get from a native app.

  3. I get notifications for sites I only intend to use once (e.g. I'm looking for an answer to a question in Google, which takes me to a forum thread on JohnsObscureWebsite.com, and so I'm just looking to get my answer and get out).

When it comes to solving the above issues, there are some options.

Point 1 (I generally get the notifications even if I have the app already):

Apple provide a specific method for linking website users to your native app, which they call Smart App Banners, which are smart enough to not show up if the user has the app installed already. Microsoft have a similar method for Windows Store apps on Windows 8. Sadly there's not really any equivalent for Google Play, Windows Phone or Blackberry World as far as I can tell.

Point 2 (I get notifications for sites that don't warrant their own apps in the first place)

Only you can answer this question, since only you know your site and the value your mobile app offering brings. As noted above, though, please outline the benefits I'll get from downloading your app if you're going to foist it upon me. If the main benefit is that I'll have your logo smiling at me every day from my home screen, that's effectively zero benefit to me. Likewise, unless your site is one of a very small range of sites that need it, installing an app to allow you to send me "special offers" is not generally benefit enough either.

If, however, it allows me to receive push notifications about something I actually care about, that would be a real benefit. Wording your app's benefits in terms of my own priorities would possibly convert me to downloading your app, e.g. imagine if you're looking at your Stack Exchange inbox on your iPhone and it said (localised to that function) "Be instantly notified about new replies and comments when they happen; download our app from the App Store". You're now offering to save me time, not take it from me.

Point 3 (I get notifications for sites I only intend to use once)

This one is probably the hardest to fix with the current Javascript solution everyone seems intent on deploying. Some ideas to fix it would be:

  • Switch to only notifying users unobtrusively (e.g. as a styled element on the page instead of as a Javascript confirmation)
  • If that is not possible, consider suppressing the alert if the referrer is a Google results page
  • Stop building native apps for sites that don't require them.

The main driver of a mobile app (or any product) is differentiation and perceived user needs. You'll need to consider the barrier of entry, as well - even if it is different and perceived as useful, you won't get adoption if it is perceived as frustrating to set up (or use). If you want to nudge users towards your app, you need to establish that the app is a good use of time (or effort).

So, to determine if you should drive (or nudge) adoption of an app, you need to answer several questions about the proposed app.

First of all, does the website lend itself to a mobile app? For example, currency exchange is much easier to access and use as an app - no web interface, just open, enter the info and ... done. So, does the app provide functionality beyond what the website can provide? Maps (navigation), flight tracking, or gift card management are great candidates for mobile app deployment. In other words, does it differentiate itself from the competition - both external alternatives, and internal (the website itself) alternatives.

Second, is the app easier to use (and setup) than the actual website. Forum management apps are a great example of this - if the user perceives value in using the app, the ability to push the app to the user is greatly increased.

Once you've established that the app is both useful and easy to use, you'll want to identify and remove the barriers to entry. Having established usefulness and ease of use satisfies some of the barriers - what is left? Do you need to promote the solution through advertising, email attachments, demonstrations? Driving adoption is highly dependent on your target audience and product.

In summary, establishing a product that meets perceived user needs is the first step. And the best way to meet perceived needs is to offer something "better". Once you've done that, you have the required foundation for nudging users.

If your app doesn't bring anything to the table, don't promote it. Just leave it as an option. There is no value added, so bothering people for something of no value is going to hurt your brand.

If it does bring value? Absolutely - and the method is up to you.


I can think of one way to nudge the user which is not that annoying. You can create a banner of your app and keep it on top of your header, so when the user lands on the page they can see you have an app, but, once they snap the header to the top (hide the banner) they have their usual experience of webpage.

Some examples were mentioned in this question: Using scrolling up from pageload to reveal additional content


Let's not forget the sample bias that can be made into these assumptions. People who download the app can also be more invested in the service and process, and by this virtue, be more likely to revisit the app. The correlation does not imply causation. It's like saying that you have more people buying ice cream in an ice-cream shop than in a restaurant.


The simple answer is, it's a sales question.

You have to give them a specific reason to close that sale.

Don't be vague, and don't offer a list: offer one specific reason:

"You're using SuperTrader on the web. Download the app version of SuperTrader to get live data" ~click~

"You're using the web version of TargetShopper. Download the app version to get the barcode scanner" ~click~

That's the way you do it.

Then, just A/B to determine which offer closes the most users.

Just talk to anyone in sales in your organisation, anyone who's really sold (cars, etc) and they'll tell you how to "make the close". Enjoy!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.